[BioC] BrainArray custom CDF repository missing in biocReposList()

Diego Diez diego10ruiz at gmail.com
Fri Apr 22 05:29:10 CEST 2011


Hi Marc,

this is great news! Thank you for your feedback on this.

Diego

On Fri, Apr 22, 2011 at 3:59 AM, Marc Carlson <mcarlson at fhcrc.org> wrote:
> Hi Diego,
>
> Your requests must have been persuasive since the maintainers at MBNI have
> just updated all of their packages again.  And so there is now a version of
> these packages available for Bioconductor 2.8 (R 2.13).
>
> biocLite() has been updated accordingly and so now you should be able to get
> these packages in the very latest release of Bioconductor.
>
>
>  Marc
>
>
> On 04/20/2011 08:54 PM, Diego Diez wrote:
>>
>> Hi Marc
>>
>> Of course, I understand (and support) the idea behind biocLite() and
>> that for the bioconductor installation to properly work it is needed
>> to ensure appropriate version of the packages to be installed. The
>> problem here is that the custom cdf packages are not - to the best of
>> my understanding- official bioconductor packages, but the repository
>> is provided in biocReposList() as a convenience. Therefore the lack of
>> support for custom cdf packages in this bioconductor version was not
>> mentioned in the release announcement. It may have been worth
>> mentioning since this may have a big impact on users.
>>
>> Obviously, removing the repository while not suitable packages exists
>> makes all the sense. However, biocReposList() also provides a
>> repository for cran, and any package present in cran will be installed
>> by biocLite() whether it is compatible with the bioconductor version
>> or not. Even there is a recent thread in the list about an old (non
>> compatible) version of the graph package present in cran being
>> installed because the official one was missing in the repositories.
>> Unfortunately I cannot think on a way to overcome this limitation.
>> That being said I myself never had a problem using biocLite() on a
>> number of year of use.
>>
>> Two additional comments:
>>
>> - The help page for biocReposList() might be a bit old since it lists
>> two repositories not presently found:
>> -----------
>> oh : URL for Bioc Omegahat package repository.  This repository
>>           contains the versions of Omegahat packages that were tested
>>           with the current Bioc release.
>>
>> li : URL for Bioc Lindsey package repository.
>> ------------
>>
>> whereas the repository "extra" is missing.
>>
>> - The cran repository is forced to be "http://cran.fhcrc.org".
>> Wouldn't be better to try to get the default CRAN repository selected
>> by the user in case that option is set?
>>
>>
>> Anyway, to finish this lengthy email just let you know that I
>> contacted the custom cdf people and they are working on providing
>> packages for bioconductor 2.8 soon!
>>
>> Diego
>>
>>
>> On Thu, Apr 21, 2011 at 1:55 AM, Marc Carlson<mcarlson at fhcrc.org>  wrote:
>>>
>>> Hi Diego,
>>>
>>> You can always throw caution to the wind and just forcibly install any
>>> tarball (if downloaded manually).  Although that would be taking matters
>>> into your own hands, and you would obviously have to live with any of the
>>> consequences that might result.
>>>
>>> But you should *not* normally be able to get these package using
>>> biocLite()
>>> if there is not an appropriate version available.  This is because
>>> correct
>>> version matching is part of what people depend on when using biocLite().
>>>  And in this case the problem is indeed because of the issue that you
>>> mentioned.  The .db packages would have been built using sources
>>> appropriate
>>> for the Bioc 2.7 and so if you installed them with Bioc 2.8, the GO and
>>> KEGG
>>> terms etc. would be out of sync with the rest of the build.
>>>
>>> It is also true that the cdf and probe packages might be ok for some
>>> things
>>> in Bioc 2.8.  But without the corresponding .db package you will not
>>> really
>>> be able to really use them fully.  And if you did use them and then tried
>>> to
>>> use those remapped probe and cdf packages with the wrong (default) .db
>>> style
>>> packages the result could be quite chaotic.  So I can't really recommend
>>> that either unless you have some very specialized usage in mind.
>>>
>>>
>>>  Marc
>>>
>>>
>>>
>>>
>>> On 04/19/2011 07:29 PM, Diego Diez wrote:
>>>>
>>>> Hi Marc
>>>>
>>>> On Wed, Apr 20, 2011 at 1:10 AM, Marc Carlson<mcarlson at fhcrc.org>
>>>>  wrote:
>>>>>
>>>>> Hi Diego,
>>>>>
>>>>> The brainarray CDFs are only missing from the 2.8 build of Bioconductor
>>>>> because there is not yet a brainarray build that matches that.  There
>>>>> is
>>>>> a
>>>>> recent brainarray build that matches Bioconductor 2.7 though, so we
>>>>> have
>>>>> of
>>>>> course provided access to that for the older 2.7 branch.
>>>>
>>>> Do you mean that the recent brainarray build was not generated using
>>>> the latest versions of the framework in Bioconductor 2.8 (like GO.db,
>>>> KEGG.db, etc.)? Because other than that I can install those packages
>>>> without problem in R2.13/Bioc2.8. I imagine this may create some
>>>> problems with unmatched entries...
>>>>
>>>>> If you are interested in access to a set of brainarray packages that
>>>>> match
>>>>> Bioconductor 2.8, then I would recommend that you contact them and let
>>>>> them
>>>>> know how much you appreciate their hard work.   ;)
>>>>
>>>> I will, since I really appreciate their hard work!
>>>>
>>>>> We will of course make such packages available as soon as it is
>>>>> possible
>>>>> to
>>>>> do so.
>>>>
>>>> Thank you for the clarification.
>>>>
>>>> Diego
>>>>
>>>>>  Marc
>>>>>
>>>>>
>>>>> On 04/18/2011 11:20 PM, Diego Diez wrote:
>>>>>>
>>>>>> Hi all,
>>>>>>
>>>>>> I noticed that the mirror for the custom CDFs available at the
>>>>>> BrainArray Microarray laboratory is missing in BioC 2.8, and wonder
>>>>>> whether this is intentional and permanent.
>>>>>>
>>>>>> Thank you,
>>>>>>
>>>>>> --
>>>>>> Diego Diez, PhD
>>>>>> Bioinformatics center,
>>>>>> Institute for Chemical Research,
>>>>>> Kyoto University.
>>>>>> Gokasho, Uji, Kyoto 611-0011 JAPAN
>>>>>> diez at kuicr.kyoto-u.ac.jp
>>>>>>
>>>>>>
>>>>>>> library(Biobase)
>>>>>>
>>>>>> Welcome to Bioconductor
>>>>>>
>>>>>>  Vignettes contain introductory material. To view, type
>>>>>>  'browseVignettes()'. To cite Bioconductor, see
>>>>>>  'citation("Biobase")' and for packages 'citation("pkgname")'.
>>>>>>
>>>>>>> biocReposList()
>>>>>>
>>>>>>                                                  bioc
>>>>>>           "http://bioconductor.org/packages/2.8/bioc"
>>>>>>                                                 aData
>>>>>> "http://bioconductor.org/packages/2.8/data/annotation"
>>>>>>                                                 eData
>>>>>> "http://bioconductor.org/packages/2.8/data/experiment"
>>>>>>                                                 extra
>>>>>>          "http://bioconductor.org/packages/2.8/extra"
>>>>>>                                                  cran
>>>>>>                               "http://cran.fhcrc.org"
>>>>>>>
>>>>>>> sessionInfo()
>>>>>>
>>>>>> R version 2.13.0 Patched (2011-04-13 r55441)
>>>>>> Platform: x86_64-apple-darwin9.8.0/x86_64 (64-bit)
>>>>>>
>>>>>> locale:
>>>>>> [1] C/UTF-8/C/C/C/C
>>>>>>
>>>>>> attached base packages:
>>>>>> [1] stats     graphics  grDevices utils     datasets  methods   base
>>>>>>
>>>>>> other attached packages:
>>>>>> [1] Biobase_2.12.1
>>>>>>
>>>>>> _______________________________________________
>>>>>> Bioconductor mailing list
>>>>>> Bioconductor at r-project.org
>>>>>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>>>>>> Search the archives:
>>>>>> http://news.gmane.org/gmane.science.biology.informatics.conductor
>>>>>
>>>>> _______________________________________________
>>>>> Bioconductor mailing list
>>>>> Bioconductor at r-project.org
>>>>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>>>>> Search the archives:
>>>>> http://news.gmane.org/gmane.science.biology.informatics.conductor
>>>>>
>>>
>
>



More information about the Bioconductor mailing list